Uniform six-day gaps hid more three-day ozone episodes than circular block gaps before selected US bird surveys in 2005-2019

We tested how missing-day order changes ozone episode detection before bird surveys. We used 797 complete 30-day windows from four US states in 2005-2019. A primary episode had three consecutive days above 70 ppb at the nearest site. We removed six days by two rules with equal daily missing probabilities. Exact calculations gave episode loss of 23.5-50.1% for uniform day subsets and 12.5-26.7% for circular blocks across two regions and two periods. The difference was 10.9-23.5 percentage points. Only 48 windows had a primary episode; the later eastern sample had one. The direction reversed for one-day events. These are finite-sample masking results, not actual failure rates or bird effects. Selected complete windows, repeated sites and rare events limit wider use. The package gives an exact counting tool, daily evidence, source hashes, checks and code.

ozone · bird surveys · missing data · episode detection · observation bias · exact counting · calendar timing